EEVblog 1386 - 295W Inverter vs 370W Solar Panel - WTF?

Why do Dave's new 370W LG solar panels use lower rated 295W Enphase microinverters?
What gives?
Inverter clipping, solar panel overrating, DC/AC ratio, solar irradiance, and microinverter underrating explained.
And a look at data from Dave's old 3kW system to predict how many days a year the new 5kW system will clip.

Buy anything through that link and Dave gets a commission at no cost to you.

Donate With Bitcoin & Other Crypto Currencies!

#Solar #MicroInverter #Rating

[arthurvlog]

Notice how your 3kW system only peaks on cloudy days right after cloud passed by. I believe it's due to overall cell temperature being lower. Clouds help to cool it down.

[bobbelman]

Belgian solar installer here, partial shading on an array of solar panels in a string should not drag down the complete output. There are 3 zones in every panel that get bridged by the built in diodes in case of partial shade on a panel.

[oliverthane]

it would not keep me awake at night, there is no sun at night to loose ... but it would keep me awake during the day ... so I guess I could also save on coffee

[jenda]

19:50 If you want to get an accurate idea of the ratio of the areas under the curve, cut out the entire peak with scissors, weigh it, then cut off the tippy top and weigh it separately. The ratio of weights will give you the ratio of the areas under the curves.

Back in the day when integrators were not available in this part of the world, we would analyze our chromatograms like this to determine the content of chemical compounds in samples. It does require a sensitive balance though.

[soundslight]

As an engineer myself, I really enjoyed your data-rich and technical video. Installers in the UK told me in string installations, they de-rate the inverter by 50% - 10kwh panels get a 5kwh inverter. Then if you have batteries, the charger has losses, batteries have 80% depth of discharge on top. Bearing in mind everything ages and degrades with time, so it seems actual usable energy is a lot less than 40% of panel rating and decreasing each passing year!

[TofuInc]

Very interesting. Slap a IQ7a or two up there and measure it over time, for science!

[funtechu]

With the solar system that I installed, we typically got 115-120% of the nominal rating during most of the day. Even after 15 years it's giving about 95% of nominal. So in my case if the inverter were only rated for 80% of the maximum, I'd estimate we would be missing 20% of the available energy over a given day on average. So depending on the cost per kilowatt it may be worth it.

I think the biggest thing is space limitations. If you are trying to extract the maximum available energy from limited space, then that's a different question if you can just throw up another panel.

[Syntax.error.]

In Europe there is less sun sure but it's also not as hot as in Australia. The hotter the panels get the fewer power they output. So I wonder how that all would count up in the end.

[nosafetyswitch]

Here in the UK, inverter undersizing of 20-30% is the standard. Some string inverter manufacturers actually allow 50% undersizing. By undersizing, I mean the DC/AC power ratio is 130% or 150%. The reasoning behind this is that you want to keep the inverter operating as close as possible to the maximum of its efficiency curve. As Dave mentions, NOCT is different than STC so you can install more rated DC power to achieve "actual" power closer to that efficency curve point.

[chocolate_squiggle]

I rent so I currently have no stake in solar, but learning all those little nuances was really interesting. So yes it'll clip, but it's not clipping every day, and the days it does clip it won't be clipping the whole day, and then the other microinverter benefits that may offset small losses anyway.... today on the internet people seem to join one camp or another so very easily and forget about all the small details. And these details can tip the scales in ways they haven't thought about. And from the comments already I see there are even more factors warranting consideration. Always refreshing to get this sort of overview on a topic, thanks.

[RomanKuechler]

I live in Switzerland; 18 Panels each with 330Wp with 18 Enphase IQ7+ Inverters (290/295 Wp). I had a clipping maybe 3 or 4 times in 2020(!) It makes really sense to slightly oversize your panels - compared to the cost of more expensive inverters.

[sameo]

“As an engineer I find that a bit offensive.” How many times have I said exactly that!

[pdath]

I think you might be missing out on generation capacity at the start and end of the day as well. The specs say it has a minimum start voltage of 33VDC. I presume this means you get zero output before this. If you had 14 panels in a string generating 33VDC each you would have 462VDC. A common string inverter would be generating a respectable output at 462VDC compared to the microinverters generating 0VDC. I imagine this effect would be greatest over the winter months.

[rjt]

The most common string system in Australia is actually 6.6 kWp on 5 kVA inverter.

Enphase micros have a longer warranty and life than most string inverters on the market, they don’t run hot internal temperatures as they run as extra low voltage (40 Voc generally).

[mrnmrn]

Dave, please note that the global solar irradiance which met stations measure is NOT the solar irradiance that your panels get, because the global solar irradiation is measured with a sensor (pyranometer) perfectly leveled horizontally. To get the exact irradiance value of your panels, you should get data from a pyranometer which is at the same azimuth and elevation angles as your panels! (This post somehow disappeared previously, so I posted it again without the link, maybe someone thought it was advertising, but it's not, I'm just happen to measure solar irradiation at work).

[dzhiurgis]

TLDW: Loosing 10% per day on 38% days equals 3.8% total loss(on ideal days). Realistically probably less than half of that, say 1-2% total loss. 
But you get same 1-2% back in improved inverted efficiency.

[sdr]

I had the same question when I received my panels (in the Netherlands) The explanation I got was that the high power inverters also have a higher loss (at 100% of the days) I checked and indeed the higher rated SMA inverter had a higher loss

[eccentriccx]

I installed a similar system in 2015 with 32xLG 310 watt panels for a 9.9kw nominal system, using enphase M250 micro inverters, and so the system peaks at about 8kw. Up here in Canada I believe this is a fairly common type of system to install. And so yes, it does clip on a perfect full on sunny day, but overall it's a very small percentage of "lost" power as you outline. Eyeball estimation would be loosing about 2 or 3kWh generation on a perfect zero cloud full summer day where it currently generates 70kWh. Perfect days are pretty rare, and I'd guess I'm only loosing 50kWh or so generation due to clipping on a given year out of a total of ~12000kWh. So although the clipping "feels" wrong, it does have very little effect, and obviously doesn't make any financial sense to pay for larger micro inverters just to capture that little bit extra power.

[filipenicoli_]

I have limited knowledge on PV systems, but looks like de-rating is quite common here in Brazil. Since the string inverters are so expensive, adding more panels is cheaper. As a result, the inverter starts clipping earlier in the morning and stops later in the afternoon, so as to use the inverter for a higher % of the time. It's not only about the energy, but the system cost.

[robertbackhaus]

Another reason to underrate - because they use more and larger switching devices, a higher rated inverter is always going to be less efficient when running at a lower power. So there will be a point, even with the same panels, where you will get more overall energy from a smaller inverter that occasionally clips, because it is slightly more efficient at other times.